[LLM] Add windows vnni binary build step (#8518)

* add windows vnni build step
* update build info
* add download command

.github/actions/llm/download-llm-binary/action.yml

@@ -14,6 +14,8 @@ runs:
         mv linux-avx2/* python/llm/llm-binary/
         mv linux-avx512/* python/llm/llm-binary/
         mv windows-avx2/* python/llm/llm-binary/
+        mv windows-avx2-vnni/* python/llm/llm-binary/
         rm -rf linux-avx2 || true
         rm -rf linux-avx512 || true
         rm -rf windows-avx2 || true
+        rm -rf windows-avx2-vnni || true

.github/workflows/llm-binary-build.yml

@@ -141,3 +141,57 @@ jobs:
           name: windows-avx2
           path: |
             build/Release
+            
+  windows-build-avx2-vnni:
+    runs-on: [self-hosted, Windows]
+    steps:
+      - name: Set access token
+        run: |
+          echo "github_access_token=$env:GITHUB_ACCESS_TOKEN" >> $env:GITHUB_ENV
+          echo "github_access_token=$env:GITHUB_ACCESS_TOKEN"
+      - uses: actions/checkout@v3
+        with:
+          repository: "intel-analytics/llm.cpp"
+          token: ${{ env.github_access_token }}
+      - name: Download w64devkit
+        shell: powershell
+        run: |
+          cd ..
+          if (-not (Test-Path ./w64devkit.zip) -or ((Get-FileHash -Path w64devkit.zip -Algorithm SHA256).Hash.ToUpper() -ne '2862F388E1720B40026F2FD95C6100A9932E3B14FB13AAC4F225A02B11E31CA9'.ToUpper())) {
+              Invoke-WebRequest -Uri https://github.com/skeeto/w64devkit/releases/download/v1.19.0/w64devkit-1.19.0.zip -OutFile ./w64devkit.zip
+          }
+          if (Test-Path ./w64devkit) { rm -r -fo w64devkit }
+          Add-Type -AssemblyName System.IO.Compression.FileSystem ; [System.IO.Compression.ZipFile]::ExtractToDirectory("$PWD/w64devkit.zip", "$PWD")
+          echo "w64devkit_path=$PWD/w64devkit/bin" >> $env:GITHUB_ENV
+          
+      - name: Build binary
+        shell: powershell
+        run: |
+          $env:Path = '${{ env.w64devkit_path }};' + $env:Path
+          make
+      - name: Move release binary
+        shell: powershell
+        run: |
+          if (Test-Path ./release) { rm -r -fo release }
+          mkdir release
+          mv build/main-bloom.exe release/main-bloom_vnni.exe
+          mv build/quantize-bloom.exe release/quantize-bloom_vnni.exe
+          mv build/libbloom.dll release/libbloom_vnni.dll
+
+          mv build/main-llama.exe release/main-llama_vnni.exe
+          mv build/quantize-llama.exe release/quantize-llama_vnni.exe
+          mv build/libllama.dll release/libllama_vnni.dll
+
+          mv build/main-gptneox.exe release/main-gptneox_vnni.exe
+          mv build/quantize-gptneox.exe release/quantize-gptneox_vnni.exe
+          mv build/libgptneox.dll release/libgptneox_vnni.dll
+
+          mv build/main-starcoder.exe release/main-starcoder_vnni.exe
+          mv build/quantize-starcoder.exe release/quantize-starcoder_vnni.exe
+          mv build/libstarcoder.dll release/libstarcoder_vnni.dll
+      - name: Archive build files
+        uses: actions/upload-artifact@v3
+        with:
+          name: windows-avx2-vnni
+          path: |
+            release

python/llm/README.md

@@ -189,11 +189,15 @@ The native code/lib in `bigdl-llm` has been built using the following tools; in
 
 | Model family | Platform | Compiler           | GLIBC |
 | ------------ | -------- | ------------------ | ----- |
-| llama        | Linux    | GCC 9.3.1          | 2.17  |
+| llama        | Linux    | GCC 11.2.1         | 2.17  |
 | llama        | Windows  | MSVC 19.36.32532.0 |       |
-| gptneox      | Linux    | GCC 9.3.1          | 2.17  |
+| llama        | Windows  | GCC 13.1.0         |       |
+| gptneox      | Linux    | GCC 11.2.1         | 2.17  |
 | gptneox      | Windows  | MSVC 19.36.32532.0 |       |
-| bloom        | Linux    | GCC 9.4.0          | 2.29  |
+| gptneox      | Windows  | GCC 13.1.0         |       |
+| bloom        | Linux    | GCC 11.2.1         | 2.29  |
 | bloom        | Windows  | MSVC 19.36.32532.0 |       |
-| starcoder    | Linux    | GCC 9.4.0          | 2.29  |
+| bloom        | Windows  | GCC 13.1.0         |       |
+| starcoder    | Linux    | GCC 11.2.1         | 2.29  |
 | starcoder    | Windows  | MSVC 19.36.32532.0 |       |
+| starcoder    | Windows  | GCC 13.1.0         |       |